Available on PyPI: https://pypi.org/project/jarvis-os/

Jarvis OS is a voice assistant designed to streamline and automate anything from daily tasks to advanced development workflows. It integrates offline speech recognition, AI-powered natural language responses, and system-level command execution. It can also function as a voice-powered command-line interface, capable of organizing, accessing, and manipulating files, as well as executing bash commands. Jarvis supports optional local customization through LM Studio.

• Offline speech recognition with Vosk

• Natural voice synthesis using Yapper TTS

• Streaming responses from local or online LLMs

  • Online default: Gemini 1.5 Flash
  • Local default: LLaMA 3.2 (1B parameters) via LM Studio
  • Customize the local model by editing the LLM endpoint URL in main.py

• Retrieval-Augmented Generation (RAG) system for web-enhanced queries

• Terminal-based interface with boot animations and dynamic loading screens

• System-level command execution, such as:

  • Opening applications
  • Performing web searches
  • Managing files
  • Shutting down system

Jarvis open vscode
Jarvis google how to bake sourdough
Jarvis shutdown
Jarvis explain quantum entanglement
Jarvis delete the read me file

It is recommended to install and run Jarvis OS inside a Python virtual environment:

python -m venv jarvis-env

• Windows

  jarvis-env\Scripts\activate

• Mac/Linux (POSIX)

  source jarvis-env/bin/activate

pip install jarvis-os

• Python version 3.8 or higher is required
• Ensure microphone input is enabled on your system
• To connect to an LLM backend (e.g., LM Studio), modify the API endpoint in main.py

• This package was initially built for Windows systems
• POSIX (Linux/Mac) support is partial and in progress

This project is licensed under the Apache License 2.0. See the LICENSE file for details.

Arvin Adeli